Jökull - 01.01.2017, Side 12

Microearthquakes on the eastern flank of Katla volcano, S-Iceland
Figure 4. Time evolution of the seismic activity on the eastern flank of Katla. Top: Registered events in the IMO
catalog between January 2001 and August 2013. Bottom: Detected events using cross correlation between June
2011 and August 2013. Two swarms are observed, one in November 2010 (top panel), the other during De-
cember 2011 (lower panel). – Tímaþróun virkninnar austan í Kötlu. Efri myndin sýnir þróunina samkvæmt
skráningum varanlegs nets Veðurstofu Íslands frá janúar 2001 til ágúst 2013. Neðri myndin sýnir þróunina
samkvæmt öllum skjálftum sem fundust með fylgnileit innan tímabundna netsins frá júní 2011 til ágúst 2013.
Tvær hrinur koma skýrt fram, ein í nóvember 2010 (efri mynd), hin í desember 2011 (neðri mynd).
the S wave (A), assuming a relationship between am-
plitude and magnitude of the form ML = c+ log(A).
The resulting magnitudes are mostly between -0.5 and
0.5 ML.
The relation between magnitudes and cumula-
tive number of events (Figure 5) follows the general
Gutenberg-Richter law (log(N) = a − bM). We es-
timated the magnitude of completeness (MC) using
the non-cumulative distribution of magnitude, where
the MC corresponds to the maximum of the distribu-
tion (see Roberts et al., 2015). MC was estimated to
be -0.4 (dashed line in Figure 5). We used the max-
imum likelihood method (Aki, 1965) to evaluate b =
1.6±0.1.
HYPOCENTRAL LOCATIONS
We were not able to locate the seismic events that
were recorded on the eastern flank before Septem-
ber 2011 with good accuracy, because the seismic
network was not dense enough and the three clos-
est and most critical seismic stations (RJU, LOD
and KKE) were not all in operation until Septem-
ber 2011. Therefore, we estimated both absolute and
relative locations only for the events detected with
cross-correlation between September 2011 and Au-
gust 2013. The absolute locations were obtained using
a probabilistic, non-linear grid-search method (Lo-
max et al., 2000), while we used a simple approach of
relative location to locate events in a local one dimen-
sional (1-D) velocity model with respect to a master
event (see Sgattoni et al., 2016b).
Absolute and Relative Time Measurements
In order to improve the absolute and relative estima-
tion of P- and S-phase arrival times, we used a cross-
correlation procedure based on the correlation of the
template event with all other events in each family.
We identified P and S phases of the template event of
the two families at each station (six in total) and se-
JÖKULL No. 67, 2017 7